artificial neural networks and deep learning
TRANSCRIPT
Artificial Neural Networks and Deep Learning- Introduction to the course -
Prof. Matteo Matteucci – [email protected]
Eng. Francesco Lattari – [email protected] also …
Prof. Giacomo Boracchi – [email protected]
Eng. Eugenio Lomurno – [email protected]
2
«Me, Myself, and I»
Matteo Matteucci, PhD
Full Professor
Dept. of Electronics, Information &
Bioengineering
Politecnico di Milano
My research interests
• Robotics & Autonomous Systems
• Machine Learning
• Pattern Recognition
• Computer Vision & Perception
Courses I teach
• Robotics (BS+MS)
• Machine Learning (MS)
• Deep Learning (MS+PhD)
• Cognitive Robotics (MS)
Enable physical and software autonomous systems to perceive, plan, and actwithout human intervention in the real world
3
Course Objectives
“The course major goal is to provide students with the theoretical
background and the practical skills to understand and use Neural
Networks, and, at the same time, become familiar and with Deep Learning
for solving complex engineering problems ... especially in vision tasks”
This is the just the 3rd edition of this course, and we,with a little help of COVID-19, just made it morechallenging by having lectures shared among twocourses. Moreover, there will be lectures you’ll like and
lectures you won’t, there’ll be topics clearly explained other not,there will be teaching styles you’ll enjoy while others will justbore you. Keep with us until the end and help us in improvingthe course so next edition will be marvelous and unforgettable!
4
A Course with Code Sharing
This course is offered to Computer Science and Engineering students
• 054307 - ARTIFICIAL NEURAL NETWORKS AND DEEP LEARNING - 5 CFU
• Prof. Matteo Matteucci, Eng. Francesco Lattari
… equivalent course for Bioengineering and Mathematical Engineering
• 056869 - ARTIFICIAL NEURAL NETWORKS AND DEEP LEARNING - 5 CFU
• Prof. Giacomo Boracchi, Eng. Eugenio Lomurno
The same teachers will teach the same topics to both classes, but you
need to be enrolled in the right course and attend the right lectures …
Bio & Mathematical Engineering students can
leave now …
… just kidding you can stay ☺
5
The Teachers
Prof. Matteo Matteucci
• Neural Networks
• Deep Learning
• Sequence Learning
Prof. Giacomo Boracchi
• Computer vision
• Deep models for vision
Francesco Lattari and Eugenio Lomurno
• Programming with Keras (python)
Official teacher, please refer to me for bureaucratic stuff!
A detailed schedule is published on the course website don’t panic!
6
A Google Calendar for you!
https://boracchi.faculty.polimi.it/teaching/AN2DLCalendar.htm
Each event includes
• Teacher
• Room / online
• Webex link to connect
• Links to video recordings
Also this link is published on the course website!
7
The Students
Students are expected to:
• To attend the proper classes according to their program
• Feel confortable with basic statistics and calculus
• Feel confortable with basic programming (python)
• Be ready to act as «guinea pigs» for this course edition
• Be curious and willing to learn ...
Students are not expected to:
• Know more than what is usually tought in basic engineerig courses
• Know already about machine learning (althought it doesn’t hurt)
• Be hyper-skilled python hackers (you’ll not need it)
• ...
???
8
Course syllabus
Introduction to Neural Network and Deep Learning
• From the Perceptron to neural networks
• Backpropagation and neural networks training
• Best practices in neural network training
• Recurrent architectures
• Autoencoders and long short-term memories
Image classification with neural networks
• Image Classification and Convolutional Neural Networks
• CNN Training Tricks and Best Practices
• CNN for Advanced Vision Tasks (Segmentation, Detection,…)
ANN and Deep Learning Coding (with Keras)
~22h lectures
~25h lectures
~14h practicals
9
Course Website and Detailed Schedule
All details and info are on the course website
https://chrome.deib.polimi.it/index.php?title=Artificial_Neural_Networks_and_Deep_Learning
How to get there?
• Goto https://chrome.deib.polimi.it
• Select “Artificial Neural Network and Deep Learning” on the left
What do you find there:
• Detailed schedule !!!
• Last minute announcements
• Slides, notes, references, lectures recordings ...
10
Lectures Schedule and Timings
Classes (there is no real distinction between lectures and exercises):
• Wednesday, 15:15 – 17:15, in T.2.1 (starts at 15:30) Team 1 (odd codice persona)
• Wednesday, 17:15 – 19:15, in T.2.1 (starts at 17:30) Team 2 (even codice persona)
• Thursday, 16:15 – 19:15, on webex (starts at 16:30 some weeks could be 2h only)
• Saturday is there just in case we cannot reach the proper number of hours
• Thursday in presence is there just in case we go fully in presence
Check the teacher who will be in class on the detailed schedule
• Connect to proper teacher webex room in case you need to attend remotely
• Use your POLIMI credentials, we will not admit external students
• Lectures will be recorded and made available afterwards
• Interaction is prioritized for in presence room students
11
Course Evaluation
Grading comprises a theoretical part and a practical part:
• Written examination covering the whole program up to 22/30
• Home project in the form of 2 coding challenges up to 08/30
• Final score will be the sum of the grades of the two 30/30
Comments and notes about the grading
• 11 points of the theoretical part will be given by Prof. Matteucci
• 11 points of the theoretical part will be given by Prof Boracchi
• 4 points for each homework challenge are given by Lattari and Lomurno
• Homework challenges are not repeated, they are just run once a year
• Laude is given to particularly active students and for optional parts
Challenges are graded based on what you do, not based on the position in the rank!
12
Course Evaluation
Grading comprises a theoretical part and a practical part:
• Written examination covering the whole program up to 22/30
• Home project in the form of 2 coding challenges up to 08/30
• Final score will be the sum of the grades of the two 30/30
Comments and notes about the grading
• 11 points of the theoretical part will be given by Prof. Matteucci
• 11 points of the theoretical part will be given by Prof Boracchi
• 4 points for each homework challenge are given by Lattari and Lomurno
• Homework challenges are not repeated, they are just run once a year
• Laude is given to particularly active students and for optional parts
Challenges are graded based on what you do, not based on the position in the rank!
13
Synergies with Other Courses
AN2DL is a course on machine learning, sinergies might exists with other
courses on the same topic, but it has been designed not to overlaps with:
• Machine Learnig: there you see classical machine learning tools, some
concepts such as generalization, overfitting, and crossvalidation might be
similar ...
• Uncertainty in Artificial Intelligence: neural networks have been removed from
this course and they have been replaced by Bayesian Networks and Graphical
Models …
• Image Analysis and Computer Vision: the feature learning part has been
removed from Image and Computer Vision , there is just a shared background
on image filtering…
• Data Mining and text Mining: does not cover neural networks and it is mostly
based on unsupervised methods
Machine Learnig: up 4-5h out of 60h (< 10%)
Uncertainty in AI: up 0h out of 60h (0%)
Image Analysis: up 2h out of 60h (< 4%)
Data Mining and Text Mining: up to 4-5h out of 60h (< 10%)
Even taking them all the overlap ends up to be at most 10h (<20%)
14
Ironing out the kinks ...
Some details have not been sorted out yet, we are working on those ..
• WeBeep Use (?)
• Projects/Competitions:
• How many people per group (?)
• When the competition will be out (?)
• Practical evaluation of challenges:
• Not doing it scores up to 0 points
• Doing it with basic tools present in class
up to 2 points (?)
• Doing it with passion and in a propositive
manner up to 4 points (?)
• Automated scoring / code plagiarism check (?)
15
Frequently Asked Question (up to now)
I cannot attend all classes, do you follow a book?You can find all covered topics on the Deep Learning book, but we are going to present the course in
a personalized manner. We suggest you to attend and follow our material then check the book to
complete your preparation. Slides will be made available as well as lecture recordings.
I am not a computer scientist (e.g., automation engineer or physics
engineer), will I be able to do the competition?
We are going to use simple libraries, we expect with basic competencies in programming you should
be able to do it autonomously at least to a minimum level
I have overlaps can I attend a different team?Teams have been defined as to balance the classes and do COVID contact tracing. It is not up to me
to check who should not be there, but it is my responsibility to check class capacity. Right now please
attend with your team, we might do some rebalancng or room exchange in case we manage.
Other questions?